| # |
Apr 11th 2019, 19:59 |
 dereuromark |
this is also how it is documented on the code doc block in public function validateUnique() |
| # |
Apr 11th 2019, 19:58 |
dereuromark |
it works for me this way |
| # |
Apr 11th 2019, 19:57 |
itmpls |
is that not how you use scope |
| # |
Apr 11th 2019, 19:57 |
itmpls |
am I missing something? |
| # |
Apr 11th 2019, 19:56 |
hollistergraham123 |
Hmmm good point |
| # |
Apr 11th 2019, 19:56 |
dereuromark |
not when you want proper validation feedback in forms |
| # |
Apr 11th 2019, 19:55 |
hollistergraham123 |
$rules->add($rules->isUnique( [‘model’, ‘foreign_key’, ‘store_id’], ‘There is already an Inventory record for this ’ )); |
| # |
Apr 11th 2019, 19:55 |
hollistergraham123 |
make it a rule I believe is a better way to do it |
| # |
Apr 11th 2019, 19:40 |
itmpls |
goal: to have a unique column varchar value with a scope of 1 column (company id). attempt: `$validator->add('title', ['isUnique' => ['rule' => ['validateUnique', ['scope' => 'company_id']], 'provider' => 'table', 'required' => true, 'message' => '..']]);` . How come this allows for multiple uniques per that company id? am I misunderstanding? |
| # |
Apr 11th 2019, 18:14 |
mike271 |
How does Garbage Collection work in a database session? Is it invoked somehow with the /usr/lib/php5/sessionclean function? |
| # |
Apr 11th 2019, 14:24 |
neothermic |
namespaces will help loads, mind |
| # |
Apr 11th 2019, 14:22 |
neothermic |
but I mean, I did fix a bunch of fundamental problems in between then and now, so there's a chance |
| # |
Apr 11th 2019, 14:21 |
neothermic |
(when we did 1.3 to 2.x, that took ~2 days to reach login successfully, so my hopes are not high) |
| # |
Apr 11th 2019, 14:21 |
admad |
i bet it won't :P |
| # |
Apr 11th 2019, 14:21 |
neothermic |
Easter's next week, might take a bit of time one evening to throw the shim + upgrade console at our codebase and see if I can even log in :P |
| # |
Apr 11th 2019, 14:21 |
admad |
it would still find problems isolated to that chapter :slightly_smiling_face: |
| # |
Apr 11th 2019, 14:20 |
neothermic |
@admad that's like trying to do a continuity edit by taking each chapter of a book in isolation :P |
| # |
Apr 11th 2019, 14:19 |
admad |
well can run them on each folder/package separtely |
| # |
Apr 11th 2019, 14:19 |
neothermic |
(which means we've had to do horrible hacks to work around ClassRegistry::init's limitations WRT Plugins) |
| # |
Apr 11th 2019, 14:19 |
admad |
hehe |
| # |
Apr 11th 2019, 14:19 |
neothermic |
(and we do that a lot) |
| # |
Apr 11th 2019, 14:18 |
neothermic |
apparently if you have two files which declare the same class and have no namespaces, they freak out. |
| # |
Apr 11th 2019, 14:18 |
admad |
though you'll have to use an older version of the tools (if they exist) for old PHP :slightly_smiling_face: |
| # |
Apr 11th 2019, 14:18 |
neothermic |
they don't like our codebase |
| # |
Apr 11th 2019, 14:18 |
admad |
@neothermic using a static analysis tool like psalm or phpstan would help you find and fix such problems. |
| # |
Apr 11th 2019, 14:16 |
neothermic |
lots of silent bugs, like doing `$foo = '';` and later `$foo[] = $something;` |
| # |
Apr 11th 2019, 14:15 |
neothermic |
thankfully I've been working on that on the side too |
| # |
Apr 11th 2019, 14:15 |
admad |
@ricksaccous |
| # |
Apr 11th 2019, 14:15 |
neothermic |
we won't be able to move to 4 directly without first updating PHP too :P |
| # |
Apr 11th 2019, 14:14 |
ricksaccous |
from what i understand anyway |
| # |
Apr 11th 2019, 14:14 |
ricksaccous |
so you can bounce to 4 :slightly_smiling_face: |
| # |
Apr 11th 2019, 14:14 |
dereuromark |
https://github.com/dereuromark/cakephp-shim/blob/2.x/Model/ShimModel.php#L312-L338 => saveFieldById() is the proper one for 2.x, and the other one can throw soft deprecation notice to help upgrading. Just have to leverage the right planning and tooling to make things easier for you. |
| # |
Apr 11th 2019, 14:14 |
ricksaccous |
4 is non-deprecated 3 |
| # |
Apr 11th 2019, 14:13 |
neothermic |
I suspect I'll get to 3 and you'll all be in 4 ;) |
| # |
Apr 11th 2019, 14:13 |
ricksaccous |
weeeee |
| # |
Apr 11th 2019, 14:13 |
ricksaccous |
join us in 3 we're waiting for you :slightly_smiling_face: |
| # |
Apr 11th 2019, 14:12 |
neothermic |
1.3 to 2.x took ~3 months and that was with the upgrade console too |
| # |
Apr 11th 2019, 14:12 |
neothermic |
this codebase is.. fun |
| # |
Apr 11th 2019, 14:12 |
dereuromark |
nah, the original one is bad, it uses a stateful ->id usually |
| # |
Apr 11th 2019, 14:11 |
admad |
or make all your table extends an `AppTable` which implements that method |
| # |
Apr 11th 2019, 14:11 |
dereuromark |
actually, if you were pro-active you could have fixed it already in 2.x to use the shim version there, wich works future-proof :slightly_smiling_face: |